Epithelial ovarian carcinoma is the leading cause of death from gynecologic cancer in the US. Approximately 75% of patients present with advanced stage disease, for which the five-year survival rate remains below 30%, whereas the five-year survival rate for stage I disease is 93%. Therefore, it is anticipated that effective methods of early detection of ovarian cancer would substantially reduce overall mortality rates for this disease. Despite much effort, there are currently no reliable procedures for the early detection of ovarian cancer available. In the past decade, a substantial amount of evidence for the occurrence of extensive alterations of DNA methylation patterns in cancer cells has accumulated. We and others have recently shown that these abnormal DNA methylation patterns can be detected in tumor-derived DNA in the serum and plasma of cancer patients. We propose to use a sophisticated automated methylation analysis technology that we have developed, called MethyLight, to screen a large panel of genes to identify markers specific for ovarian cancer, compared to non-neoplastic ovarian tissue. This will allow the correlation of methylated markers in ovarian tumors with clinicopathological features, and with response to chemotherapy and overall survival. Subsequently, we can use the information obtained in this screen of tissue samples to develop markers for the detection of ovarian tumor-derived DNA in the serum of patients with existing or recurring disease. Finally, we propose to use the most promising biomarkers to evaluate the capacity to detect preclinical relapse of disease, as a function of time before clinical diagnosis of relapse.